News for hackers. Stuff that matters.

A comment on GPT-5. Parent comment

Re:GPT-5 by anon7000 on Thursday August 07, 2025 @03:03PM (#44828927)
True. At a minimum, as long as LLMs don't include some kind of more strict representation of the world, they will fail in a lot of tasks. Hallucinations -- responding with a prediction that doesn't make any sense in the context of the response -- are still a big problem. Because LLMs never really develop rules about the world.

For example, while you can get it to predict good chess moves if you train it on enough chess games, it can't really constrain itself to the rules of chess. (https://garymarcus.substack.com/p/generative-ais-crippling-a...)

[ Parent | on HN ]
Re:GPT-5 by math_dandy on Thursday August 07, 2025 @04:06PM (#44829692)
Two schools of thought here. One posits that models need to have a strict "symbolic" representation of the world explicitly built in by their designers before they will be able to approach human levels of ability, adaptability and reliability. The other thinks that models approaching human levels of ability, adaptability, and reliability will constitute evidence for the emergence of strict "symbolic" representations.
[ Parent | on HN ]
Re:GPT-5 by sharemywin on Thursday August 07, 2025 @04:09PM (#44829734)
but you could easily build a verifier and if it's not valid have it create a new move until it finds one.
[ Parent | on HN ]